I’ve been listening to quite a bit of EconTalk lately, which has been a really fascinating and illuminating experience. I do find the show is dragged down a bit, though, by a twinge of self-righteousness on the part of both Russ Roberts and many of his guests regarding the virtues of the free market that leads to occasional congratulatism as well as the snarking at straw men. That aside, though, Roberts is very willing to invite and engage guests of all ideological stripes in good faith and I recommend the show to anyone interested in economics.

One thread I’ve noticed after listening to a few dozen episodes, though, is the obvious disdain in which Roberts and many other economists hold the minimum wage. The best single example is this discussion with Mike Munger, but it recurs in many episodes. This surprises me especially since he and Taylor Hanson, in what is a truly fascinating discussion, actually make a very compelling implicit case in favor of the minimum wage when discussing The Singularity:

These would just be cheaper employees. Because? Why would they be cheaper? Actually, Ricardo got this right back in 1820. Published a paper on robots substituting for people. Simple calculation: found that when a person and a machine are direct substitutes, then the wage the person gets can’t be any higher than the cost to rent the machine. When the machine gets cheaper, the wage falls. If the world economy is dominated by employees for which it’s easy to make more of them, then the wages have to fall at the rental price of the machine. Under a scenario where people make nearly as many machines as possible and still not lose money. More of a zero-profit constraint.

Did you catch that? No? OK, here’s Roberts interviewing Adam Davidson talking about Standard Motor Products:

Davidson: On the cost side, his main cost is metal. That’s a bigger cost for him than labor, and metal is a globally traded commodity; he’s not big enough to strike some deal with Alcoa to get preferential pricing. So, labor isn’t the only cost he can cut, but it’s one of the main costs he can cut. Russ: If he can find a way to substitute machines. So, as you mention in the article, you don’t replace all workers with machines because sometimes workers are cheaper. Same issue of China versus the United States–sometimes you are going to have coexisting workers with the machines; but you are going to be looking for ways, when it’s possible, to do it cheaper, and whichever way that is, which tends to be as labor gets more expensive over time, machinery is going to sometimes dominate. Davidson: Right. In the case of Maddie, the young woman that I focused on, so she makes $26,000; with benefits and everything let’s call it $30,000 a year. And they did the math, and a robotic arm could do what she does more quickly and more precisely. But it would cost $100,000. And their metric is: Any capital investment needs to pay for itself within two years. And so Maddie makes $30,000; over two years $60,000, the net present value of which would be less; so an upfront expense of $100,000, it’s clear Maddie gets to keep her job. Russ: But as the price of the machine comes down. Davidson: Or if there is more demand for fuel injectors and they add a second or a third shift, the mathematics change. Quickly. And there certainly are robotics manufacturers out there trying to cut their costs. And Maddie isn’t in a position to cut her costs very much. This is a non-union plant by the way, but $13 seems to be about the floor for a manufacturer like this.

So connect all this:

  •          There is a point at which, economically, it makes sense to replace a human worker with a machine.
  •          That point is determined by the relative costs of the human and the machine.

So logically, if you use a price floor to artificially increase the cost of human labor, you have created more scenarios when a machine becomes economical.

But so what? More people are replaced by machines? How is that a benefit? The way to understand this is to look at it dynamically. Just building a single machine in a “snapshot” view of the economy and replacing a human with it does no good – but if you look at the economy progress over time what you are in fact doing is accelerating the pace of innovation and technology-based efficiency and productivity gains by creating more opportunities for economies of scale. Take somebody working at McDonald’s whose job may be threatened by automated touchscreen ordering. If they cost their employer, arbitrarily, $25K a year in total compensation, and the machine costs $45K, by the Standard Motors logic McDonald’s should buy the machine. But in a world with no minimum maybe McDonald’s decides “hey, we could threaten this guy’s job and get him to take $20K a year instead” and it works because the guy needs the job and somebody else might be willing to take it at $20K and he accepts. Or he doesn’t but McDonald’s does find somebody else who will do the same job for $20K. What you have now is a world where you have some combination of “McDonald’s is more profitable” and “Big Macs are a little cheaper and there’s more consumer surplus” but you don’t have a world where McDonald’s placed a huge order for touchscreen ordering stations. But thanks to the minimum wage in our world, they did! And now because of economies of scale maybe the systems are a little cheaper so Wendy’s gets them too. And soon everybody has touchscreen ordering systems far sooner than they would have in the other world where all they do is bid down wages.

Now for this to really work to improve human welfare you have to have education keep up with innovation, so that as there are fewer jobs in “taking orders for Value Meals” and more jobs in “manufacturing and maintaining touchscreen ordering systems” you’re not supply-constrained in terms of labor. But if you can make this work you have a really big win – you may have sacrificed some short-term gains in terms of efficiency but you’ve essentially delayed the inevitable world where these kinds of technology replaces human labor and there are fewer degrading jobs and more rewarding jobs and everyone is better off. In fact, this is very much a society-wide savings where we make ourselves a little worse-off in the present but we are creating more capital – machines, factories, increased knowledge and innovation – that makes us increasingly wealthy at each point in the future.

This argument doesn’t mean the “correct” minimum wage should be $100/hr, but it does suggest that the minimum wage plays a more useful function in promoting economic growth than many conservatives or classically-trained economists might suggest, in addition to the traditional progressive arguments about individual rights, welfare, and dignity.